Cell lines are critical reagents for much of cancer research. One might intuitively imagine that cancer cells, with their in vivo selective growth advantage, would be relatively easy to establish in culture. However, generating cancer cell lines is surprisingly difficult. Athymic ("nude") and severe combined immunodeficient (SCID) mice are valuable tools that support human cancer cell growth. When the cancer is harvested for in vitro cell line production, fibroblasts and other stromal cells commonly overgrow the culture dish and in most cases prevent isolation of the malignant cells. The study of invasion and metastasis (hallmarks of cancer) is similarly hindered by the inability to recover small numbers of cancer cells that have completed the process. To circumvent these problems, we are generating immunodeficient mice that are hprt defective, and have already constructed a nude hprt-null mouse. Similar to other immunodeficient mice, they will support growth of many xenografted cancers. During tumor growth, mouse cells replace the human stromal cells. When we wish to recover the human cells, we will eliminate the mouse stromal cells by growing the culture in the classic selective media, HAT. In this proposal, we plan to: generate SCID hprt-null mice, demonstrate superiority of the biochemically defective mice for cell line production, and demonstrate enhanced recovery of cells from spontaneous metastases. Specific Aim #1: Generate biochemically selectable immunodefective mice designed for human cancer cell recovery. Specifically, generate SCID hprt-null mice. Specific Aim #2: Document superiority of the hprt-null mice over standard nude and SCID mice. Specifically, document mouse stromal cell replacement and isolate pure populations of human cancer cell lines from xenografts established from patient-derived primary cancers (breast, prostate, lung, pancreatic and glial cancers). Compare recovery with and without stromal cell supplementation. Specific Aim #3: Document utility in recovering small numbers of metastatic cancer cells. Specifically, determine recovery efficiency from spiked samples, and isolate organ-specific metastases from experimental metastases and from spontaneous metastases after orthotopic implantation. Mice will be deposited at Jackson Labs for distribution to investigators at non-profit institutions, including the NIH.